Database Research Summaries2017 Investigation and demonstration of the potential for unmanned aerial systems to detect and quantify stressors in soybean production fields
The focus of this research is to inform producers about advantages and disadvantages of using unmanned aerial vehicle (UAV) technology for farming operations.
Determine the potential utility of UAV technology to identify specific stressors in fields in order to determine the feasibility of investment in the technology.
- Researchers flew approximately 70 UAV missions over soybean research plots. Plots covered research on herbivory, herbicide injury, and herbicide efficacy. The missions were conducted with multiple sensor payloads (spectral), at multiple altitudes (spatial), and with high frequency from planting to harvest (temporal) to evaluate the necessary resolutions for detecting target problems.
- Researchers used a texture analysis algorithm to evaluate the potential for automation of herbivory damage at the levels used in the study and was applied to R1 and R5 soybean.
- At R1, the algorithm is able detect the 50% clipped row and classify it as one class; the 25% falls in between the clipped and non-clipped, but the algorithm cannot detect a 10% clip. At R5, the algorithm is able to detect the 50% clipped row, but it also selected quite a few other blocks, not within the 50% clipped row. The results indicate that near infrared image information would have assisted with automated detection of herbivory damage.
- UAVs suffer from poor battery life. Lower altitude flights often take longer to cover the same area; therefore, it is in the best interest of the operator to conduct UAV missions at the altitude that provides the spatial resolution necessary for detection. The results indicate that herbivory damage was visible even at the 300 ft. altitude with the Phantom.
- Researchers captured and documented prickly sida populations in sample areas within research fields at Stoneville and Starkville. They were able to see weeds with still images and video footage from the Phantom UAV. Weeds appeared more yellow than soybean in Starkville, but the “greenness” of soybean can vary even by variety. This means that color alone may not be a reliable indicator of weeds. Weeds were less detectable when small, indicating that weeds might be beyond the treatment window before reliably detectable.
There is potential to detect production problems with soybean with a UAV. This type of technology allows farmers to detect stressors quicker and more efficiently.
For more information about this research project, please visit the National Soybean Checkoff Research Database.
Funded in part by the soybean checkoff.